| gnarface | Alverstone: well, that's the old, old way of doing it, but someone wanted dependency-aware booting added to sysvinit and then added it to debian a full year before systemd. despite this, the major justification for systemd was "sysvinit doesn't support dependency-aware parallel-booting" :( | 00:04 |
|---|---|---|
| gnarface | so now there's a minor conflict in behavior, but the way it seems to work is that LSB headers take precedence over the /etc/rc*.d/S* (start) symlinks, but /etc/rc*.d/K* (stop) symlinks take precedence over the LSB headers instead | 00:05 |
| gnarface | i guess this was all to eliminate the overhead of distro maintainers having to repeatedly re-index and agree to indexing orders on banks of symlinks in unrelated projects | 00:06 |
| gnarface | it's occasionally been a pain in the ass for me, but less often than you'd think | 00:07 |
| gnarface | in theory you could get what you want done by just moving cryptdisks-early to S00 or S01 or something | 00:09 |
| gnarface | but that'll not necessarily also work as expected in sysvinit | 00:09 |
| Alverstone | It was just a pretense. systemd's overtake was a planned and coordinated. On subject, since I use runit I don't see any sysvinit specific utilities that do the actual parsing and dependency resolution... for some reason. /etc/init.d/rc seems to do something, but I don't really understand what. | 00:11 |
| Alverstone | there are some .depend* files, don't understand them either | 00:12 |
| gnarface | eh, i'm vague on the implementation details but i think that whatever isn't in /etc/init.d/rc, /etc/init.d/rc.local, or /etc/init.d/rcS is in the /sbin/init binary itself | 00:14 |
| gnarface | it's not such a large and complicated thing as people made it out to be | 00:15 |
| gnarface | oh, there's also some configuration in /etc/inittab but i think nothing really directly related to dependency tracking | 00:16 |
| gnarface | hmm, though now that i look that up, stackexchange says they're parsed by /etc/insserv | 00:19 |
| gnarface | (and chkconfig on redhat) | 00:19 |
| gnarface | dunno if that's true but it sounds legit | 00:19 |
| Alverstone | update-rc.d | 00:33 |
| Alverstone | good, now I can sleep peacefully | 00:33 |
| Alverstone | finally | 00:33 |
| gnarface | uh... i don't think that one does anything about LSB headers though... just the symlinks... | 00:38 |
| rwp | gnarface, You have it! But let me tune that up a little. inserv reads the LSB headers and writes out new S* and K* symlinks based upon them. And most importantly writes out /etc/init.d/.depend.boot and .start and .stop Makefile make compatible dependency information. | 01:43 |
| rwp | Then at boot time those are used in dependency order in parallel like "make -j#" and I think at least was at one time actually make but don't know for sure now. At boot you will see the message "Using makefile-style concurrent boot in runlevel 2." If those files are not present then it falls back to the S* ordering. | 01:43 |
| rwp | Before then maintainers could never coordinate on numbers. And they could never grasp the problem of breaking circular dependencies at boot time. So then they converted to the inserv tool which complains when there are circular dependencies. People sometimes ignore the complaints. | 01:45 |
| rwp | The classic circular dependency loop is that everything wants syslog online to log messages including DNS servers. Everything wants DNS online to look up names including system log daemons. Everything wants network attached storing online including addressing them by name. | 01:47 |
| rwp | You can see the types of loops that are created. And which must be broken by some method that a human must choose. But people never want to address bootstrapping issues like that and dig in their heels and force the problem into the initramfs. | 01:48 |
| gnarface | rwp: thanks, that helps a lot! i will probably forget all that by the next time it comes up but it's good to know someone around here knows the details | 02:44 |
| rwp | My turn for a question. Yesterday one of my VMs at Digital Ocean started producing this error. https://paste.debian.net/1335460/ | 02:51 |
| rwp | E: /usr/share/initramfs-tools/hooks/growroot failed with return 1. | 02:52 |
| gnarface | not sure i've seen that before, i've only got two guesses: maybe you're out of space on a partition or maybe /boot is marked read-only | 02:52 |
| rwp | Anyone know the problem with this? Or an easy way to debug it? There is just no information emitted. | 02:52 |
| rwp | Good question. All in one partition. /dev/vda1 ext4 20G 17G 1.7G 92% / | 02:53 |
| gnarface | or wait, not just read-only, i think if it's mounted with "noexec" i've seen kernel choking too | 02:53 |
| rwp | Hard to believe it would need more than 1.7G of space to create the new initramfs but that's possible. I wonder what I can prune there. | 02:54 |
| gnarface | wait but it's ext4 so doesn't it have a reserved percentage as well? | 02:54 |
| gnarface | if it's reserving 5% or something that makes it more likely you could have run into the limit | 02:55 |
| gnarface | the kernel modules are almost 400 MB here | 02:55 |
| rwp | That 1.7G will include the reserved 5% percentage. | 02:56 |
| rwp | I found 2G that was wasted junk that I could delete immediately. Up to 3.8G free now. Same error. Can't believe it needs that much space. | 02:56 |
| gnarface | what package is /usr/share/initramfs-tools/hooks/growroot in? i don't have that here | 02:58 |
| rwp | This system is part of a project which is mostly idle now. It's now my turn to say that nothing has changed. :-) But clearly upgrades have been continuing automatically upgrading. And then for some reason this one daedalus system can't make initramfs on vmlinuz-6.1.0-26-amd64 or vmlinuz-6.1.0-27-amd64 | 02:59 |
| rwp | Ah... "update-initramfs -u" with -v prints copious information. Chasing down the rabbit hole of it. | 03:01 |
| gnarface | cloud-initramfs-growroot - automatically resize the root partition on first boot | 03:02 |
| gnarface | part of this package maybe? | 03:02 |
| rwp | Yes. Definitely part of that package. | 03:03 |
| rwp | Thinking about purging that package in order to avoid this error. | 03:03 |
| gnarface | it seems like you shouldn't need to grow anything on an already-created VM disk image | 03:04 |
| rwp | But I am looking down through those scripts now. Ran manually and am getting this "/usr/share/initramfs-tools/hooks/growroot: 204: verbose: parameter not set" But it might be because I am running it manually without any particular env variables set which it might need. | 03:04 |
| rwp | Also clearly /usr/share/initramfs-tools/hooks/growroot is shorter than 204 lines so that confuses me. | 03:05 |
| gnarface | well one would think it should be smart enough to realize the partition is already 100% of the available disk space, but i dunno | 03:05 |
| rwp | I also don't see where it actually grows the partition either. It just copies some programs around in the style of UsrMerge. | 03:06 |
| rwp | That package was recently upgraded too. Would count as something that changed recently. | 03:10 |
| rwp | This system has not yet been UsrMerged. This seems to be an error resulting from that situation. | 03:16 |
| rwp | YES! It's a fallout from UsrMerge and this system just hadn't gotten it yet. Because it's mostly an idle thing now. | 03:24 |
| rwp | The /usr/share/initramfs-tools/hooks/growroot script is trying to copy /usr/bin/grep /bin but on an unmerged system grep lives at /bin/grep so there is no /usr/bin/grep to copy and the script fails. | 03:24 |
| rwp | And it is the same for three utils in that script all of grep, rm, sed have the same issue. | 03:25 |
| gnarface | wait wait | 03:26 |
| gnarface | so a usrmerge script... | 03:26 |
| gnarface | is failing because it's unusermerging something? | 03:26 |
| gnarface | there seems to be a lot of irony in that situation | 03:26 |
| rwp | To work around the problem I copied grep, rm, sed from /bin to /usr/bin so that the /usr/share/initramfs-tools/hooks/growroot can find them and copy them from /usr/bin to /bin successfully and continue. | 03:27 |
| * gnarface facepalms | 03:27 | |
| gnarface | this is jumping the shark level shit here | 03:27 |
| rwp | I then re-ran the apt-get upgrade allowing it to trigger the update-initramfs and it completed successfully. | 03:27 |
| rwp | I don't really know what the author was trying to do there but take a browse of /usr/share/initramfs-tools/hooks/growroot and tell me why they are doing that set of copy actions. | 03:28 |
| gnarface | i feel like this is going to come up again | 03:28 |
| gnarface | i don't have that file here though and i'm not gonna install it | 03:29 |
| rwp | I script the upgrade and it runs every day on these systems. This is a Devuan 5 Daedalus Stable system and it hit this problem yesterday. I just got to it today. | 03:29 |
| rwp | Let me paste it. https://paste.debian.net/plain/1335462 | 03:31 |
| rwp | I didn't install cloud-initramfs-growroot but it came as part of a standard Debian install on Digital Ocean and then upgraded to Devuan later, it's been around a while. But cloud-initramfs-growroot is just a standard package in the repository. | 03:32 |
| rwp | I will bet that people will hit this on random cloud hosted systems. | 03:32 |
| gnarface | hmm, yep, pretty blatantly copies them right up front | 03:32 |
| gnarface | comment but no explanation... | 03:32 |
| rwp | And now you know as much as I do about it! | 03:32 |
| rwp | The root cause of the failure was basically copy_exec /usr/bin/grep /bin can't work because there is no /usr/bin/grep, and the same for /usr/bin/rm and /usr/bin/sed. | 03:33 |
| rwp | I did a workaround by "cp /bin/{grep,rm,sed} /usr/bin/" to ensure that they did exist there so that this hook script could do it's own copy in reverse and succeed. | 03:34 |
| rwp | This is only ever going to be a problem on non-usrmerged systems. Which this one is but only because of neglect. | 03:34 |
| rwp | And this happened in Stable this last week for some reason. The file dates on the files in the cloud-initramfs-growroot package are dated Aug 20 so recent change. That should not happen in a Stable release but it escaped through. | 03:35 |
| rwp | Probably a super easy workaround would be to purge the package. And I would have done that except I wanted to get to root cause of the problem first. | 03:36 |
| rwp | Stuff here IRL. BBIAB. | 03:36 |
| onefang | Got an email from my ISP telling me about an emergency outage between "23:00 and 05:00". Then goes on to say "work will be performed during daylight hours." If they meant between 11 AM and 5 PM, which would make more sense, then it would be off already. Sooo, I could drop out suddenly for hours, or that could happen later tonight. | 03:50 |
| onefang | So much for my plans to get stuck into apt-panopticon improvements today. If my network is going away for hours at some random time, testing network stuff isn't a good plan. | 03:51 |
| onefang | Time to eat. No network required. B-) | 03:52 |
| * systemdlete lumbers on with the nut-2.8.2 effort for daedalus... | 05:37 | |
| systemdlete | good news/bad news. I got the patches working by carefully updating them. Got past that. Then hit the renamed tmpfiles file problem and fixed that by updating the configure.ac file. | 05:40 |
| systemdlete | Got farther along, but now hitting doc file issues. At the step where it tries to generate the docs, it complains it cannot find the "UPGRADING" file. | 05:41 |
| systemdlete | oddly, this seems to come from the same rule in the Makefile; it is able to make the others, but not that one. All have *.adoc source files. | 05:41 |
| systemdlete | dh_installdocs: error: Cannot find (any matches for) "UPGRADING" (tried in ., debian/tmp) | 05:46 |
| systemdlete | it seems to be calling or referencing a make rule in debian/rules I looked at the lines indicated in the error messages, but I don't see an explicit dependency in either of those rules for that file. | 05:49 |
| systemdlete | in fact, drilling down a bit, looks like it fails on | 05:52 |
| systemdlete | dh_installdocs -A ./README ./NEWS ./TODO ./AUTHORS | 05:52 |
| systemdlete | (where is it getting the idea of "UPGRADING" from I wonder...) | 05:53 |
| systemdlete | I see... UPGRADING is referenced in the named files | 05:58 |
| systemdlete | so it looks like UPGRADING.adoc did not get processed and put where the process expects it. OK... | 06:01 |
| systemdlete | (see, this is why I'd rather die than go through all this...) | 06:01 |
| onefang | Please don't dlete systemdlete. | 06:07 |
| systemdlete | I'm gonna cheat here and just copy the .adoc file to the target. I don't plan on using it anyway... | 06:12 |
| systemdlete | whoever upgraded this apparently forgot to upgrade the build process for generating the upgrading info... | 06:12 |
| systemdlete | (maybe they meant to keep the secrets of upgrading to themselves, who knows) | 06:13 |
| systemdlete | had to cheat a few more of these adoc files... | 06:21 |
| systemdlete | I will report this deficiency to the package author to have them fix their doc make steps | 06:22 |
| morenonatural | hey, y'all ... probably not the best place to ask, but dunno where to look for this: I have an old usb gamepad (HID era) and I want to use it as an XInput (basically any PC gamepad nowadays) | 15:44 |
| morenonatural | probably "writing a module" is a solution, but if a) there's already a driver (but registers as a non-XInput device) b) re-writing a XInput gamepad (template available?) ... | 15:46 |
| morenonatural | I would like to use a) or b) if available, minimize the work to do | 15:46 |
| gnarface | i was under the impression that there was a hardware requirement so older gamepads can't do it | 15:48 |
| gnarface | but the first thing that comes to mind as examples would be the existing modules... you can get the kernel source package and look at them | 15:48 |
| gnarface | ...there might be a userspace component required too? i'm foggy on it... | 15:49 |
| freem | what's HID already? | 15:49 |
| morenonatural | Human Input Device | 15:49 |
| gnarface | i know steam will still support non-xinput gamepads anyway though, for whatever that's worth | 15:49 |
| freem | ah... like, a mouse, a keyboard, etc | 15:49 |
| morenonatural | if you know XInput, it's the same concept | 15:49 |
| gnarface | you don't necessarily need xinput to use your gamepad, mind you | 15:49 |
| morenonatural | gnarface, even in linux distros? | 15:50 |
| freem | yeah | 15:50 |
| gnarface | yea | 15:50 |
| morenonatural | ohh, nice | 15:50 |
| morenonatural | will check that out | 15:50 |
| morenonatural | it's not necessary, but it definitely makes things easier | 15:50 |
| freem | I mean, keyboards and mouse on linux can work even in the TTY after all | 15:50 |
| morenonatural | not all games support non-XInput gamepads... lesser do as time goes by | 15:51 |
| freem | I think xinput and consort are more about making things supported by gtk/qt | 15:51 |
| freem | are not most games built on top of SDL? | 15:51 |
| morenonatural | I'm thinking games played through wine | 15:51 |
| freem | still | 15:51 |
| freem | SDL is not just for hobyists | 15:52 |
| freem | well, I don't know too much on this | 15:52 |
| morenonatural | I can check that out too... my restrictions are "time to sit and code", but I still have "time to read about stuff on my phone" (I got kids and I take em to the park) | 15:53 |
| morenonatural | * I got kids and have a culture load that provides me with 10min-ish spans available to read spread through the day | 15:54 |
| gnarface | a surprising amount of games are actually built on mono, even on linux "native" titles | 15:54 |
| freem | yes, since Unity and now Godot target C# as primary or secondary target | 15:55 |
| gnarface | even stuff older than unity was for linux | 15:55 |
| freem | I have bought 2 games based on Unity. Shitty performances, I have good reasons to suspect the engine's the root. So godot might really be a blessing | 15:55 |
| gnarface | i think | 15:55 |
| freem | next time I evaluate a game, i'll check the engine, and if unity, then nope. | 15:56 |
| gnarface | in my experience the unreal engine does a lot better | 15:56 |
| freem | (but that's offtopic) | 15:56 |
| freem | well, UE was a strong competitor to quake | 15:56 |
| djph | Unity could use some love; but I've not seen it be /terrible/ | 15:56 |
| freem | talking about FPS... you can't waste cycles nor network frames :) | 15:57 |
| freem | nah | 15:57 |
| freem | unity merits despise, and apparently they manage to buy it with their latest director or whatever, giving nice publicity to godot | 15:57 |
| djph | oh yeah, I don't play FPS (or at least not multiplayer) | 15:57 |
| freem | and it's a full win for us: even if the game is non-foss, the engine being foss means we can fix perf problems :) | 15:58 |
| freem | or try. | 15:58 |
| freem | for certain can report them, at least | 15:58 |
| freem | Unity's director did us a great favor trying to grab more money :) | 15:59 |
| freem | regarding the original issue, I never owned... err, ok, I did once, but it was a serial one... an usb gamepad or alike, so I have no idea how they are handled by the kernel | 16:00 |
| gnarface | some of them are handled by more generic hid modules, some of them have specific modules | 16:01 |
| freem | I did wrote a software directly using libinput (great lib, btw! neat API, trivial to use, it was actually fun to code with it) and framebuffers, for work, but is all | 16:01 |
| freem | SDL failed at handling the Xorg-less (and wayland-less) setup, it's why I had to resort to NiH | 16:02 |
| freem | we were using a touchscreen on top of a screen. The stuff basically appeared exactly as if it was mouse | 16:02 |
| freem | previous attempt by a colleague to use SDL, which failed, colleague tried (and to some success, have to say, but it was buggy) to directly read the device's file in /dev IIRC | 16:03 |
| freem | that does not really helps with the issue though | 16:04 |
| freem | morenonatural: maybe you could lend us some info about the device? | 16:04 |
| onefang | I was about to ask which device are we talking about. | 16:06 |
| onefang | Long ago I had my PS3 USB controller working with LInux and Android. Long ago being when PS3 was new. Don't recall details. | 16:07 |
| morenonatural | https://www.amazon.com/Buffalo-iBuffalo-Classic-USB-Gamepad/dp/B002B9XB0E | 16:07 |
| AlexLikeRock | i got controller ps3 generic | 16:08 |
| morenonatural | I believe that's actually newer... I don't remember "Android" mentioned at any time | 16:08 |
| AlexLikeRock | help me to make usefull | 16:08 |
| freem | not until you rename to Alex Kidd | 16:09 |
| morenonatural | not to be confused with the wireless, either | 16:09 |
| AlexLikeRock | I was stuck in the step, that if it detects it as HID, but not as a joystick | 16:09 |
| AlexLikeRock | yes, mi controll are both , wire and wireless | 16:10 |
| morenonatural | ah, this sounds like it https://www.reddit.com/r/snes/comments/18i6vz/buffalo_japans_retrostyle_8_button_usb_gamepad_is/ | 16:10 |
| AlexLikeRock | half detects it as HDD with bluetooth | 16:11 |
| AlexLikeRock | with wire, never | 16:11 |
| morenonatural | I can plug it and give y'all more info... lemme know what command I have to run | 16:11 |
| freem | err... and usb snes controller?!? | 16:11 |
| gnarface | js-test | 16:11 |
| gnarface | plug it in and run dmesg | 16:12 |
| morenonatural | AlexLikeRock, https://www.linuxnest.com/how-to-use-ps3-controller-on-linux-a-comprehensive-guide/ not useful? | 16:12 |
| freem | morenonatural: dmesg for a start, then lsusb, likely | 16:12 |
| gnarface | maybe look to see if any files in /dev/input look newly created | 16:12 |
| morenonatural | I remember "sixaxis" back-and-forth from some years back | 16:12 |
| gnarface | ps3 is a special case i think, but the generic usb gamepad will probably "just work" | 16:13 |
| gnarface | (as a legacy gamepad /dev/input/js0 or something, likely) | 16:13 |
| AlexLikeRock | thanks morenonatural | 16:13 |
| AlexLikeRock | that what i nee , just work | 16:13 |
| AlexLikeRock | hehehe | 16:13 |
| morenonatural | dmesg: https://paste.debian.net/1335508/ | 16:14 |
| * freem just realised... is JS for reverted Sad Joke? | 16:14 | |
| morenonatural | wtf, `lsusb command not found` | 16:15 |
| freem | "apt-get install usbutils" IIRC | 16:15 |
| onefang | sixaxis is a name I recall from my playing around with the PS3 controller. Fairly sure the Android driver for it included that name. | 16:15 |
| freem | or just check result of "apt-file searc bin/lsusb" (this is a very useful tool) | 16:15 |
| freem | search* (ofc) | 16:15 |
| morenonatural | lsusb (I think): Bus 003 Device 003: ID 0583:2060 Padix Co., Ltd (Rockfire) 2-axis 8-button gamepad | 16:16 |
| onefang | I should pull that PS3 gamepad out and try again over the weekend. Hook it up to my MIDI setup like my other joysticks. | 16:17 |
| AlexLikeRock | i found my controller :-D | 16:19 |
| onefang | Mine will be in the box with the rest of the PS3, and I know where that box is. | 16:20 |
| AlexLikeRock | its this one : https://www.powersystemonline.com/MCO-1277983606-gamepad-unitec-doubleshock-pc-ps3-inalambrico-_JM | 16:21 |
| gnarface | morenonatural: it recognizes it, using the "hid-generic" kernel module | 16:22 |
| gnarface | ...maybe only as /dev/hidraw3, but i'm not clear on that, check in /dev/input for new files too | 16:24 |
| AlexLikeRock | my ps3 : found ! : Bus 001 Device 002: ID 054c:0268 Sony Corp. Batoh Device / PlayStation 3 Controller | 16:24 |
| AlexLikeRock | jstest-gtk : not found! | 16:25 |
| gnarface | morenonatural: run "ls -lht /dev/input/" and check the files at the top, are any dated for today, when you plugged it in? | 16:25 |
| AlexLikeRock | http://paste.debian.net/1335510/ | 16:26 |
| gnarface | AlexLikeRock: do you need bluetooth for this? | 16:27 |
| AlexLikeRock | not now , first by USB wire , please | 16:27 |
| AlexLikeRock | http://paste.debian.net/1335511/ | 16:27 |
| gnarface | i think there's a sixaxis package... | 16:27 |
| gnarface | or at least was... | 16:27 |
| AlexLikeRock | apt-get install sixpair sixaxis | 16:28 |
| AlexLikeRock | Reading package lists... Done | 16:28 |
| AlexLikeRock | Building dependency tree... Done | 16:28 |
| AlexLikeRock | Reading state information... Done | 16:28 |
| AlexLikeRock | E: Unable to locate package sixpair | 16:28 |
| AlexLikeRock | E: Unable to locate package sixaxis | 16:28 |
| AlexLikeRock | :-S | 16:28 |
| gnarface | well, if you just rebooted that's not useful, they're all the same time... | 16:28 |
| gnarface | might not be there yet though, due to those errors: | 16:28 |
| gnarface | "failed to retrieve feature report 0xf2 with the Sixaxis MAC address" | 16:28 |
| gnarface | "failed to claim input" | 16:29 |
| gnarface | could be /dev/hidraw0 though | 16:29 |
| morenonatural | gnarface, I see /dev/input/js0 | 16:29 |
| AlexLikeRock | ls /dev/hidraw0 | 16:30 |
| AlexLikeRock | ls: cannot access '/dev/hidraw0': No such file or directory | 16:30 |
| gnarface | morenonatural: there you go, anything that will use the legacy interface will recognize it then anyway. xinput... probably not. | 16:30 |
| gnarface | morenonatural: (assuming your user has read permission on that device, that is) | 16:31 |
| gnarface | morenonatural: though i do wonder if it would be possible to somehow make a custom driver that could do it without any special hardware... | 16:33 |
| gnarface | that's a noble goal and i think you should keep looking into that | 16:34 |
| gnarface | AlexLikeRock: maybe it needs to pair on bluetooth first for some reason? | 16:35 |
| AlexLikeRock | how to do ? | 16:36 |
| gnarface | need to find a bluetooth pairing program | 16:37 |
| AlexLikeRock | morenonatural, do you now it ? | 16:37 |
| gnarface | AlexLikeRock: his works by default because it uses a different driver. can you see if the hid-playstation driver is loaded for you? | 16:38 |
| gnarface | not sure that's the right one but i see it there | 16:38 |
| gnarface | i see one suggestion to just install xboxdrv and run that to emulate a xbox360 controller with it... | 16:40 |
| gnarface | Steam might work with it too using just its own internal support... | 16:41 |
| AlexLikeRock | yes, i read about it some time ago | 16:41 |
| gnarface | looks like you might need sixpair from a 3rd party source | 16:42 |
| gnarface | i'm not seeing it in the distro... | 16:42 |
| AlexLikeRock | thanks gnarface | 16:43 |
| AlexLikeRock | https://github.com/conradev/Sixpair | 16:43 |
| gnarface | i seem to see mention of several different projects doing this, not sure which is best or most current | 16:44 |
| AlexLikeRock | dam! | 16:44 |
| AlexLikeRock | "Pair a PS3 Sixaxis controller to your iPhone " | 16:44 |
| AlexLikeRock | could work at GNU ? | 16:44 |
| gnarface | i see some people saying it should "just work" over usb even, but clearly it didn't | 16:45 |
| gnarface | so, i'd say try this | 16:45 |
| AlexLikeRock | its fine to me | 16:45 |
| AlexLikeRock | just USB | 16:45 |
| gnarface | did you install joystick and jstest-gtk? | 16:45 |
| AlexLikeRock | yes | 16:46 |
| AlexLikeRock | its basic :-> for my others controlllers | 16:46 |
| gnarface | hmm, debian wiki says it should work out of the box now too | 16:47 |
| gnarface | weird i wonder what you're missing, what release are you on? | 16:47 |
| AlexLikeRock | same error : failed to retrieve feature report 0xf2 with the Sixaxis MAC address | 16:47 |
| AlexLikeRock | 16:47 | |
| gnarface | ? https://wiki.debian.org/Gamepad#Xbox_and_PlayStation_controllers | 16:47 |
| gnarface | says it should work | 16:47 |
| gnarface | do you have Steam? | 16:48 |
| AlexLikeRock | i dont like STEAM | 16:48 |
| gnarface | https://help.ubuntu.com/community/Sixaxis | 16:48 |
| gnarface | well i found this for ubuntu | 16:48 |
| AlexLikeRock | i used : emulator : PS , N64 , Znes , FCEUX ,etc | 16:48 |
| gnarface | ...but i wouldn't recommend using a ppa from ubuntu | 16:48 |
| gnarface | some of the setup instructions might still help though | 16:49 |
| AlexLikeRock | not found: apt-get install sixad | 16:49 |
| onefang | Just stumbled across kodi-eventclients-ps3, so seems there is some PS3 support buried in our repos. | 16:49 |
| AlexLikeRock | could addd repository from ubuntu to my devuan :? | 16:50 |
| AlexLikeRock | apt-add-repository ppa:falk-t-j/qtsixa | 16:50 |
| gnarface | please don't do it | 16:50 |
| gnarface | you could screw up your install | 16:50 |
| AlexLikeRock | so... not work to me : https://help.ubuntu.com/community/Sixaxis | 16:50 |
| gnarface | well, but you could find the source and build it though, probably | 16:50 |
| gnarface | then follow the rest of the instructions | 16:51 |
| gnarface | it's just that you don't want a package referring to ubuntu dependency versions screwing up your next system update | 16:51 |
| gnarface | it would be fine to build a package against the debian versions, probably | 16:51 |
| AlexLikeRock | X_X | 16:51 |
| gnarface | it can't be that complicated of a piece of software | 16:52 |
| gnarface | don't die!! | 16:52 |
| AlexLikeRock | hahahahah | 16:52 |
| onefang | Damn. Everything is in the Playstation 3 box, except the gamepad. lol | 17:20 |
| onefang | It'll be in some random crate somewhere in my storage room. | 17:21 |
| AlexLikeRock | steam-devices | 17:39 |
| AlexLikeRock | This package provides udev rules for various Steam-related hardware devices such as the Steam Controller (gamepad) and the HTC Vive (virtual reality headset). | 17:39 |
| AlexLikeRock | :-> | 17:39 |
| AlexLikeRock | still : fail! .:-( | 17:41 |
| n4dir | I don't need this often. I start the jack audioserver with cadence or qjackctl, both gui tools. If i log out from the GUI, using fluxbox, jack stops. Both have a setting to keep jack running, but it doesn't work. | 19:00 |
| n4dir | I assume i miss something a DE has, or a more complete installation in general. Any ideas? | 19:01 |
| rwp | n4dir, I don't really know about jack (a general statement!) but can you script starting it in a ~/.xsessionrc file which is automatically sourced by the Desktop Environment upon start. And also include it in your fluxbox start (.xinitrc file probably) and then it will always be started by either X way? | 19:21 |
| rwp | In my ~/.xinitrc file I put ". $HOME/.xsessionrc" and then put things I want shared in both in that file. Works for me. For other things anyway. | 19:22 |
| n4dir | rwp: oh, now i remember. Also at least cadence has jack to be started as soon i start cadence, and that doesn't work either | 19:22 |
| n4dir | rwp: this isn't a real problem. It is more about understanding what might be responsible for that | 19:23 |
| n4dir | fluxbox uses .fluxbox/startup (not that it would matter) | 19:23 |
| AlexLikeRock | . | 19:25 |
| n4dir | say stuff like dbus and what not, i really have no clue what they do. (dbus is installed, just giving an example) | 19:25 |
| AlexLikeRock | pkaction | grep mount | 19:25 |
| AlexLikeRock | Error getting authority: Error initializing authority: Error calling StartServiceByName for org.freedesktop.PolicyKit1: Launch helper exited with unknown return code 1 | 19:25 |
| rwp | n4dir, When you start an X Window session running fluxbox, what command do you use to start it? I might assume either "startx" or "xinit" as those are the typical ways to start a window manager session as opposed to a Desktop Environment session. A DE session is most often started through an X Display Manager xdm such as lightdm or slim is currently the default installed one. | 19:38 |
| n4dir | it is startx | 19:38 |
| n4dir | i do start DEs via startx too, might be uncommon though | 19:38 |
| rwp | Using startx will look for a ~/.xinitrc file and then that does not exist fall through to /etc/X11/xinit/xinitrc. | 19:39 |
| rwp | The /etc/X11/xinit/xinitrc sources /etc/X11/Xsession which sources ~/.xsessionrc file. So putting something in ~/.xsessionrc file should be included in a startx startup sequence. | 19:40 |
| rwp | Note that ~/.xinitrc if it exists must include everything that you want. In other words if you just put one thing there and don't do the rest of /etc/X11/Xsession then those won't happen and it will appear broken such as no window manager started. | 19:41 |
| n4dir | rwp: i wouldn't even now what to autostart | 19:41 |
| rwp | By interface design one usually has many things that usually happens there such as setting up their keyboard, loading ~/.Xresources, setting the background, "stuff", and then the last line is exec'ing the window manager. When the window manager exits the X session exits. | 19:43 |
| n4dir | yeah, but say mocp or screen run on in the background. And according to cadence and qjackctl, jack should do that too | 19:44 |
| rwp | My last line is "exec i3" (though just "i3" works too but just leaves one excess sh process around). | 19:44 |
| n4dir | now that i think about it: how come mocp keeps playing music if jack stops? It gets more weird. | 19:45 |
| rwp | Programs that are started and run as daemons should continue running. | 19:45 |
| rwp | Programs that connect to the X session should terminate when the X session terminates. | 19:45 |
| n4dir | ok, about that i was wrong. If i exit X, jack stops and as expected mocp stops to give audio. What a relief | 19:50 |
| n4dir | same for xfce, btw. | 19:51 |
| rwp | Anything connected to the X session should terminate when X terminates. But if something runs as a daemon not associated with X then it should continue since it is not associated with X. | 19:51 |
| n4dir | Yes. | 19:52 |
| n4dir | The question is how to make jack keep running | 19:52 |
| fsmithred | start jackd when you log in? | 19:52 |
| n4dir | fsmithred: if i do it from cli, yeah, then that works, but my old script to start jack doesn't work anymore, no idea why. | 19:53 |
| n4dir | so i use either cadence or qjackctl. And both have said "problem" (not really a problem, as i log out like once per month) | 19:53 |
| n4dir | both have a setting to autostart jack when they are started, both doesn't work. | 19:54 |
| fsmithred | oh | 19:54 |
| n4dir | at least qjack has a setting to keep jack runnning even if, but that doesn't work either | 19:54 |
| fsmithred | I've used that setting in the past (and probably now, too) | 19:54 |
| n4dir | on the laptop i assumed it works with a DE, but on this PC it doesn't work with xfce either | 19:54 |
| fsmithred | nothing relevant in the message window in qjackctl? | 19:55 |
| n4dir | i will look there. Here cadence is installed. | 19:55 |
| n4dir | Thu Nov 14 19:49:55 2024: WARNING: Disconnect message was received from D-Bus. | 19:56 |
| n4dir | Thu Nov 14 19:49:55 2024: WARNING: Disconnect message was received from D-Bus. | 19:56 |
| n4dir | damn, double post. | 19:56 |
| n4dir | perhaps it really is dbus, of which i know no nothing, is involved | 19:57 |
| rwp | D-bus is a component of the message-passing operating system that has been created around D-bus. The idea in a message-passing OS is that that there will be various management process daemons running and no program does anything themselves but send a command message to the daemon responsible to have it do it instead. | 20:02 |
| n4dir | so might it or might it not be involved? | 20:05 |
| AlexLikeRock | init 6 | 20:09 |
| rwp | n4dir, It might be involved. I don't know. Plausible either way. (Don't type in init 6 as that will reboot your system.) | 20:18 |
| n4dir | i mean, if i have to log out i have to restart terminal-emulator, reattach to screen, restart the web-browser, so i guess restarting jack isn't that important | 20:20 |
| n4dir | reopen all tabs in web-browser, re-login, what not | 20:20 |
| rwp | I have never run jack so I don't know what makes it so difficult to run. | 20:28 |
| AlexLikeRock | help | 20:44 |
| AlexLikeRock | i broke dbus | 20:44 |
| AlexLikeRock | dbus-monitor | 20:44 |
| AlexLikeRock | Failed to open connection to session bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. | 20:44 |
| AlexLikeRock | . | 20:44 |
| AlexLikeRock | dolphin (file manager) : | 20:45 |
| AlexLikeRock | An error occurred while accessing '3.7 GiB Removable Media', the system responded: You are not authorized to perform this operation: Not authorized to perform operation (polkit authority not available and caller is not uid 0) | 20:45 |
| AlexLikeRock | . | 20:47 |
| AlexLikeRock | /devices/pci0000:00/0000:00:14.0/usb2/2-2/2-2:1.0/host4/target4:0:0/4:0:0:0/block/sdb/sdb1 (block) | 20:47 |
| AlexLikeRock | UDEV [2199.774451] bind /devices/pci0000:00/0000:00:14.0/usb2/2-2/2-2:1.0/host4/target4:0:0/4:0:0:0 (scsi) | 20:47 |
| AlexLikeRock | do i need https://wiki.gentoo.org/wiki/Polkit#Configuration ? | 20:53 |
| AlexLikeRock | my iss | 20:53 |
| AlexLikeRock | my system they not have | 20:53 |
| AlexLikeRock | pgrep -a polkit | 21:13 |
| AlexLikeRock | no are running | 21:14 |
| AlexLikeRock | how to start? | 21:14 |
| AlexLikeRock | id polkitd | 23:58 |
| AlexLikeRock | uid=110(polkitd) gid=119(polkitd) groups=119(polkitd) | 23:58 |
| AlexLikeRock | service polkit status | 23:58 |
| AlexLikeRock | polkit: unrecognized service | 23:58 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!